ENGLISH ACCENTS

by D. Crystal

There has been so EUchwritten on the subject of 'English accents'that it might seem presUI!1ptuousto add to t..""estockpile, and inconceiv­able that there could be anything new to say. After all, any summ~rschool or lecture course worth its salt gives a period or two over tothis issue; choice of accent provides a perennial, and consequently afamiliar problem, both for the teacher and the being taught; there isplen~ of detailed descriptive information available (for example, inthe second half of A.C. Gimson's Introduction tn the Pronunciation ofEnglish); and the linguist's and phonetici~'l's'~~ttitudes-to the phenomenonof accent are nowadays well-1mownarid appreciated. Certainly, with allthis verbiage, it would be silly and unnecessary to repeat or even sUJ;DIlla-,rise all that has been said about such funda-mental issues as the difference between'accent' ~d 'dialect', or about the variouspopular misconceptions as to what accents areand how they should be evaluated (misguidedcriteria of 'correctness', and so on). (Any­one not familiar with "t.'1.i·sbackground shouldread David Abercrombie' excellent presenta­tion of the situation in Problems and Princi­El~s_withoutdelay). But 'd.~·;PTteall this in­terest and attention, there is surprisinglyan enormous .amount which has not been said.Accents are deceptively complicated thingsstill.

It is the very _ fa:.;uliari ~ of '-the ;ho­tion of accent which makes it so dangerous,of course. Our abili ty to forrr,ulate a fewgeneralisations about it, couched in vaguelyscientific terms, produces a disease calledcomplacency, which obscures the real and deep-rooted difficul ties in~lved.This is fD.OStundesirable, as there are still beny irlportant factors whichhave not. been systematically considered in relation to accent (e~g. thenotion of s~les of speech), there are distinctions which have not beenclearly made, though they need to be (e.g. that between voice-quali~ andaccent), and there is no theoretical framework which explains and inter­relates all theforIi1al features which enter into the definition of anaccent, and the functional factors which relate ti1e use of differentaccents to different si tua tions. Manyof the probler:1s and confusions whichteachers cif English overseas have can be reduced - possibly even removed ­if one begfus.t.ostudy such olTQssions•

.Anytheory of a0cent must show the relevance and na ture of both ..linguistic and si tuatirmal factors. CDmprehensive, accurate discussion ofaccents in English, or any language, wiil not ta!ce place until at leastthe following influential variables have been identified and. discriminated.Variables in the situatinn in which we speak are of five types, which weDay.label ~rsona.1., $i~2~c, region~, E.I!1!>..§,. and ~~stic. Features inthe forD or substance nf 2...11 accent, which do different jobs, are of four

4 4

kinds: the basic ~hone~_i£ system) the ~~op~ system, non-segmental

features, and ~n-li~i~tic features. Different problems arise at eachpoint in these lists of relevant factors ,:ve shall briefly discuss each

in turn, beginning with the situational factors, as these will probablybe the most familiar.

The personal factor in our discussion of accent is usually referredto under the general heading of voice-quality, This is very oftenconfused with a,ccent, though it should-be)~p-t clearly apart. While we

speak, apart from the actual message we are trying to put across, we alsocommunicate information of a quite different kind, operating at an entire~

different level. This is information about our pe+sonalities. Whenever

we speak, we make known our identity to the outside world; there are

features of everyone's voice which allow others to recognise an individual

without seeing him. These features are difficult to pin down precisely,

but they clearly exist, and they are very different from the rest of our

utterance. Voice-quality is a relatively permanent feature of our speech;

it only alters with age or physiological change (as when we develop a

hoarse throat, for instance). Mimics (people who deliberately imitate

other people's voice-quality) are the exception in society rather than the

rule. Normally, people can do nothing about their voice-quality, nor do

they usually want to o~ange it -- unless they have some professionalinterest in mimicry or acting. (A sil:ilar situation exists in writing:

a person's handwriting is the factor which allows us to recognise anyone

for who he is, and we only alter this on very exceptional occasions).

The main contrast between voice-quality and accent is that,the

latter is a more general phenomenon. It can be used to refer to the

totality of sonic features a person has (including his voice-quality), as

in 'T ~ like his accent' (i.e, way of speaking); or it can be used in amore restricted sense to refer to those non-idiosyncratic features of a

person's pronun ciation .~ that is, those sounds. which are· shared by a

number of other people and whichoinform us that someone comes.from a

particular region or social group. Voice-quality tells us who someone

is; accent tells us where he is fro:'.l,either diachronically (:Le.

temporally), regionally, or soci&lly.

Let us now examine the iL:;>lications of this last sentence. First,

what do we mean when we say that accerit conveys diachronic information

as to origin? Essentially, we mean that a perso~'s accent can tell us

the general period in history in which he was born and brought up. Thisdoes not cOme to the fore very often, of crlUrse, but we do notice

diachronic'ccntrasts of this type whenever we become aware of phoneticdifferencesbe~Neen generations of people. "d1en we talk of a 'conservative'

pronunciation, for example, we usually mean a pronunciation characteristio

of older people. Younger generations ,in],:ng::Landare frequently singled

out by their distinctive 'advanced' pronunciation, which is primarily dueto the influence of American English.

Such changes, as is usual in comparative diale ctology, primarily

affect the vowel sounds, and a useful survey of the most noticeable

changes from the traditional realisation of Received Pronunciation can befound in an article called 'Phonetic Change and the RP Vowel System',

written by A.C. Gimson, and published in In Honour of Daniel Jones

(Longmans, 1964, pp.131-6). He points t';-~t118~e nature of the finalelements of closing diphthongs, for example =~ a tendency to mono­

phthongise eI, OY, a1, a.tr and 01 , so that pairs of words like ~Land ~, ~~ and ,~!~ become homophonous; and describes how thecentring diphthongs 1El, £. El , 08 and V8 are also changing, particularly intheir relation to the pure vowels, J , e, re and 0: (for example, 08 nuw

only exists in the ·speech of older speakers of RP, or of people directly

influenced by them ~ in all other cases, it has been levelled with ~: ,

so that ~ and ~ are homophones). These along with a number of

,ther cha.l1ges,r;rc.:..;:e c::"'es.r evidence of the existence of stable phonetic

differences oe~e~ ±e ge..'1erations. Manyof these differences, affectthe pronunciation of ~Cividual words, as opposed tf) being generaltendencies; for example, changes in the stress pattern of small groups ofwords, such as '~troversy', which is nowmore frequently 'con~ersy'.'l'he catalogu:,i.ngof these changes is a long and difficult task, but a greatdeal of information about then( will be incorporated into the revisededi tion of Dar:,iel Janes' English. Pronouncinei £i9..t12E~r"y',which is due outin the near future.

Diachronic information, in view of the very slow rate of languagecl1ange, is 'not by any means as obvious as the other kinds discussed below.If we had some means of observing longer stretches of speech history, then

presumably we would be able to note diachronic differences very clearly,in much the same way as we can readily associate different styles of hand­:rri ting with different historical periods. Until the advent of time­machines, or the discovery of primi ti ve gramophonerecords, however.• we areforced to re~ or- scholarly reconstruction of the ~lder states of a language,based on writing and other clues. This is always appr0ximate, but we do nowhave a reasonably good idea of the sound system current in ChaucerIs andShakespeare's times and since.

Temporal changes in accent are of course often due to the influenceof regional or social factors - for example, the develnpment of pcip,recordsfrom America leads to changes in pronunciation amongst the younger generationnf British speakers on~. But regional and social factors (the eecond beingdivisible into claSS and styli~~i£ factors) are of importance in themselves,and we must now discuss each in turn.

Regional information, first, is probab~ tbe most usual thing weread into anyone's accent, and there should theref0re be no need to labourthe point: an accent can identify where in the w0rld or a country a personis from. But what is often forgotten is that there are two quite differentkinds of regional accent in English these days, which we might labelintranaticnal and international. The intranational accents should not be newto anyone, being the subject-matter of traditional dialectology, andreferring to those accents vn1ich can be distinguished vnthin a given coun~.The international accents have n0t been so well studied, however. These arethe national accents which have developed due tn the spread of Englishthroughout the world. Despite the differences which exist between Welsh,Lancashire and LondonEnglish, there is a cow~onbond which makes themdistinctively British, as opposed to American and Australian. And there isa similar bond for the different dialects of American English. Standardaceounts of the phonetics and phrmology of these mo.jnr regional dialectshave yet to be written, in the majority of cases. 4e must remember that.quite a few areas are involved - apart frnm Britain and the U.S.A., there isCanada, Australia, NewZealand, South Africa, the1dest Indies, and manyoB1er areas where English acts as a second language, Only the first twohave sufficient global status and influence to present the teacher of Englishwith any selectir)llal problems, however. Weshall discuss this further below.

Therein one further thing to be said aboutintranational regionalaccents, hmvever, and this has importa.nt implicatinns for pedagogy-. This isto note the development of ~ local reginnal dialects (~rgn accents). Itis sometimes saH that English regional dialects are dying. This is a grossoversimplificati0n. It is true that manyof the old, rural dialects arebeing spoken by fewer and fewer people - this is being madevery clear bysuch surveys as the one currently being carried on in Leeds (the first two

lumes of which have already been published); but these are being replaced,so to say, by the growth of new urban dialects, such as those of Liverpool

Bircingham.

?he acce~ts as~~~~~~s~ -:~~ ttese cities are t~emselves variegated:~re a..:-e 5~e=eJ.:- !.."..=;:'': 5.=:" =.-::'~: 50r::etiilles verging nn the unintelligible

6

save nnly to theini tiated, whic~1.arc !:~.::.~ytt,= ':'-~;::~ ')f the foreignteacher; and there are more general, st~c_3.:-,="'5.::;~ f'lr;:.:s, intelligible toeveryone, which are. These latter acceLts ~~=~ s ray composite forms,nuxtures of the local accent and. F2. I~ ~s ~~~ ~~cult to see howthese 'mndified RP' accents develnped. 3scause c~~es are cultural,ecnnomic and educational centres, they cr~t~~ ~~luential people, whndo not wish to change the accent with which tney are brought up, and whosee no need to. Those daysaro gone when one had tn conform to theeducated London accent to get sncial prefernent. But sn too are thosedays gone whenoi ties were isnlated c()I:'.munities. Due to the rapiddevelopment of inter-city cnrsmnication, and the availability of ,ocal'~lasst:1edia, influential penple come into contact with other influentialpeople more frequently than ever before; and the result of contactbetween friendly social groups is inevitably partial standardisation.Canaequently, we find very clear tendencies fnr speakers nf urbandialects, to conform to more generally'used kinds of accent - and in viewof the traditional prestige atkched to RP, they usually stray in thatdirection.

These days, then, the concept of a single, pure, monolithic,homogeneousRP; always mak7-.11f1.:.seof the same set of phonetic and phono­logical contrasts, is a myth. There are manykinds of p~, mixtures ofthe traditional RP and a regional accent. Sometimes the traces, ~fregional influence are very slight: I have heard speakers where the ~nlyclifferencebetween their speech a:1 ~ traditional RP was the use of afor iin words like cup. In such 82,'3eS,indeed, it is impossible to definethe speaker's place of origin, as a nlliTlberof regional accents couldhave caused this change. In a-cher co-Des,the 'modified RP' might be veryolearly identifiable vdth a particular area - as in the Yorkshire-basedspeech of the British Prime Niinister j for example. 'rhe moral for theforeign learner of English, theY', i~.> to be aware of the compleJdty of,the situati'on, and try to get convE'rsant with therrain regional modifi­cations of RP; and vvhile he will pri:;::"ablylearn tbe regionally mostneutral form himself (Le. the GTaditional form), he must not go awayvd.th the impression that all deviations from this .<Jtandardare equallybad! The moral for the phonetician is that he should desoribe, nota single accent, but the ~ange of accents which constitute educated orgener&lly acceptable speech.

Wecan see from all this talk about eduCation and pres ti ge thatthe theoretical distinction behveen regional and class accent is mUchblurred these days, at least as far as English is concerned. (It isclearer in some other languages, such as Welsh). For English, accentcan give someinformation about class differences - identi~ whoa personis in terms of a social scale of some kind - but this information is not·very reliable. To begin with, no-one is really very sure about howmanysooial classes there are anyway: the old triad of working, middle and .upper has very little scientific standing, largelybecaus.e of the problemof' defining the boundaries between the supposed classes. Moreover, evenif one grants this threefold distinction as a working hypothesis, it hasbeen shown that it is impossible to predict most people's class merelyfrom their accent, and vice versa.

Blit while granting these difficulties of obtaining precise' inform­ation, it is nonetheless the case that there are strongeintuitive grounds forpostulating -that accents in at lea.st British English carry some overtonesof social background, or cla.ss. Objective evidence for this is scattere~and we await the first sociolinguistic surveys of this question eagerly,but it does exist. Satirical reviews and comedians make frequent use ofaccents to point a joke about social differentiation; and a phoneticiancolleague of mine can elicit a 'sir' out of most bus-conductors by askingfor his ticket using a highly conservative form of :RP! It is on accountof a multiplici ty of stories such as these that 0ne would be forced toincorporate class factors into any theory of accent.

_ •• 8 l.as~ s:.~s. _':'C.::.E.1. ::actor I want to discuss I have labelled

t:v:.istic;by th:..s= s~ ~=-:.:r.eanthe association of a particular accent,or 'tone of voice' (as :.tis often called in this context) with a

_articular kind of social situation, usually an occupation. ,One would,consciously or otherwise, wake use of a different accent from what one

.ould normally use in everyday conversation, as soon as one began a

articular task. The accent associated with preachers, barristers,

tr_dertakers, television advertisers, politicians speaking in public, and

o on, would all be cases in point. All of us have at least two accents~hich we make use of regularly - one for informal occasions, and one for

~ormal or professional occasions. Many public figures accumulate quite

a range of accents to meet various situations. The foreign learner' can,

initially, make do with one - a relatively informal variety; but he must

remember that he cannot use this with equal appropriateness in all.,English-speaking situations, even if he is fluent in it. This is no~'a

~ estion of linguistic correctness, of course, but of linguistic manners.

- ~se an informal accent (or dialect) on a formal occasion produces asimilar kind of criticism to using informal clothing on a formal occasion.

-ie may not ]jke this situation, and foreigners are usually forgiven for

temporary breaches of linguistic etiquette; bu'l:;the situation exists,

and if you are trying to get as near perfection in your English as possible,then you should try to master it •.

It should be clear from this brief survey that selectional problemsfor the teacher will depend on his assessment of the relevance of each of

the above situational variables. Voice-quali~y can usually be discounted

in 'the field of teaching English as a foreign language, though I have come

across cases of courses designed to give students an accent modelled on

at of SirWinston Churchill or Bing Crosby. (Doubtless the Beatles are

ei..l1gdeified as a phonetic model in ~_Jme dark corner of Rio at the

:::oment!) Elocution classes for native speakers of English certainly use

is technique quite often~ Stylistic factors are also unimportant,

a?art from the question of formality just discussed, though again I have

heard of courses teaching a restricted kind of English accent to teachers

ting to use it only at conferences. Class fD.ctors should also provide

Y:o problem, as I doubt vvhetl1eranyone would ever want to teach anything

other than a form of the educated, prestige accent.

The only problems, in fact, are connected with the diachronic and

regional factors - particularly the international sense of the latter.

As far as diachrony is concerned, it is probably best to teach the more

conservative forms of accent, partly because their status is more well~ .established, partly because these are the forms which are most satisfac:"

torily described in the textbooks, which in the nature of things mustalways be a little or a lot out of date.

In the case of international regional accents, the only real problem

s the question of deciding between a General 1\.T\1.ericanpronunciation or

D, and here it is ~-linguistic faotors which will dictate the answer._.e will teach that accent which the student is most likely to come into

::tect within his English-speaking life, and this vrill be largely

~ecided by which country his own is most strongly influenced by, either

_ li ticaily, culturally, ecanornically, educa tionally, or in any other

~~:r. If it is impossible to decide this questiiJn (it rarely is ) then theea~~er has no choice: he must be arbitrary, and take one and stick to itir he must teach both. And we all know which s0lution all sane teachers

~ould prefer! But in such a dilemma,arbitrariness is not to be criticised.

~cat I would crilicise would be the abs.:;nceof ~L attempt to make studentsa•.•are of the ,most important differences between major regional accents, for

-his is sonethingwhich cnuld easily be brought in at the more advanced

::eges 0:' a course ~ or even earlier, 'if the students were in such a

-':'y-:::::' ~G>sh-spe<>k-''''g C0121try that the differences between, say, British:::..:::' ;...::.s=-=-c~ 7."n-...:lc. =r .:-cet:.~e:::.sel\esor: :..~eirattention daily.

We can thus be :'l5x:..-:=-= :..::: ~~ ::;';~- -:-..:.::Z to~ard3 choosing an

accent to teach; but, having chose:-d .r;:; ::...~t i,:'", ~..:::. -:e precise as to the

nature of the formal features cor:s...:.~~~~ :":-.'-=::::'5 ::"6adsus on to the '

second part of our theory of accer:";:,.rc.i.:::-. =-~:- :::::;-r us how to distinguish

between the components of an accer.";:-::.:..,::.~-.~'.::--='~a::t jobs. There is no

space to go into any of ~hese in ce:~~ ::.=~=) ~~~ ~~&t =ust be pointed outis that there is a clear progress:"c~ :~~==~ ~= ~o'~ aspects specified

above -- a definable ordering of pr~c~_":'55, ~ :5~-8 of the criteria of

in telligibili ty and overall acceptat~ -~ -::-.The four aspects were: ~5 :~5:"Cs:·s'e= of segmental phonemes,

the system of allophoncs, ~1e syste~ ~~ ~c~-5eg=E2val features, and other

vocal features ofa non-linguistic nat;urE. 1o..s ~ar as any question of theintelligibili ty of one's English is cor.cerned, one has to be able to master

the phonemic contrasts before anything else; and this, in the'nature ofthin,gs, means being able to master at least one allophone per phoneme

(for phonemes are abstractions which can adopt any number of shapes de­

pending on the phonetic context - we cannot speak n:w phoneme directly,but only one of its possible shapes, i.e. allophones). And (to carry the

argument a stage further), as allophones can only really be clearly ex-.plained by describing ill the allophones of a given phoneme more or lessat once, (try explaining what'clear l'is without any reference to 'dark Itfor example) to master the phonemes is in effect to master the majority ofthe allophones also. So the first two components, separable in theory, are

really one in practice: they must be mastered if intelligible English is to

result -- and intelligibility is one of the prerequisit~s of acceptability.I think it is probably true that teachers would like to cover

the entire range of allophones' in any course, all duly graded and drilled.

I find that people are more uncertain as to exactly how much non-segmental

vocal effect they should instil into their pupils. Now by non-segm~~taleffect I basically mean all those linguistic contrasts which are caused by

variations in the pitch, loudness or speed of the voice - such as the

intonation patterns, and the stress. Here the teacher must distinguishcarefully between those parts of intonation and stress which cause clear­

cut differen~es in meaning, of a grammatical type, and those which ,do not.The former - for example, the difference between a restrictive and a non­

restrictive relative clause, which in speech is entirely made through non­

segmental features - is clearly something that has got to be taught, as it

affects the basic intelligibility of an utterance. The latter are not sobasic, though they too may affect intelligibility, in various degrees.

These aspects of intonation are what is normally referred to

under the general heading of the 'attitudinal t function of intonation,and few teachers try to teach this 'systematically, but hope (beyond hope! )

that students will pick this up, as they themselves did. This is not the

place to argue the' case for and against the teachability of intonational

attitudes. I personally think this will be perfectly possible as soon asmore research has been done. The point I want to make here is that

mistakes of an attitudinal sort in using intonation are not grievous sins

against intelligibility, usually. ,If one is know~ to be a pleasant

foreigner, a,ttitudinal' f.§:..L1]£ pas· are normally tolerated - or at least onecan get one's message across, even if the response ca~ries with it a queerlook. Consequently, thisdoes'no.t pose a very urgent pedagogical problem.

In other words, inability to master the entire range ofEnglish's intonational patterns wi,ll not seriously affect our ability to

understand or communicate in English., Similarly, inability to master the

non-linguistic features of an accent complete~ - voice-quality featureswhich do not affect the meaning of C.LT utterances, but which do contributetowards the national identification of an individuaJ., e.g. American

nasality - will not affect our communicational ability either. Such

inabilities should only give us an inferiority complex if we seek perfection

in mastering an accent - andbyper~ection here I mean that a foreignershotlld not be recognised as such by. the linguistically-naive maJ:lin the

street (the criterinn is not, heaven, help us, a phonetician!) '.):he ultimatetribute a foreigner can have is to be asked 'You're not from round here,

'7 9

are you? \1hic.hpart 0:'" ~ie country are you from?' Such perfeet;i.ol'\ isquite attainable; but even if you yourself have been recognised as comingfroI:\ abroad, take heart! - nine people out of ten in the South of Englandthink that (according to their accent, at any rate) people from Neweastleshould cnmefrom abroad too!

+ -+- + + + -+- -+- + RE G ION AL+

The British Council vall be able to assist teachers whoare accepted for the School with bursaries, to cover a largepart of their costs ~ the exact amount will vary according to

e distance they have to travel.

The January School for the Southern Regi~n will al~osteertainly be held in Curi tiba, and the State of parana has verykindly offered its support and assistance. we are very gratefulto the State Education Secretary for the enthusiasm with whichhe has received our plans •

In the meantine however, we can say this much: that theSchools will be concerned entirely with the English language andways of teaching it. Vie shall keep the wnrk as practical aspossible, and the emphasis will be on methodology rather thanlinguistics. A great deal of the time will be spent in smallseminar groups, and that is (me reason whywe are limiting thenumbers to 50 or 60 at the most. The staff vrill be mainlyBritish.

+++++

++++++++

+

+

+ +++++++++++++++++++++++++++

SCHOOLS + + + + -+- + +SUMMER

Wehope that each of the two Schools will cater for about50 or 60 active teachers. These will be selected mainly fromSta te Schonls in as llany parts of the Regi0n as pnss:i:ble =- al­though Cultura teachers will certainly be eligible to apply,especially if they also work in State schools. Weshall be send­ing rouhd in the near future supplies of application forms, andwith these will comefurther details of the Southern Regioncourse.

Following the success of the International SumoerSchoolin Rio last January, the British Council h2.Sdecided to hold twoRegional SummerSchools for Brazilian teachers of English, onefor Southern Brazil in January 1968, and one for Northern Brazilfor July 1968. Weare not quite sure I'lhere the dividing linecomes: but perhaps teachers from Rio might consider themselvesas ..Southerners for the occasion, and those from Brasilia asNortherners.

++

~

~

•

+++++

+++++

•

•

++

++

+

...

...

..

..

·+

...

If any of our readers wish to, they are very welcome to.l.i te to ne at once, and I vall send the.r:tthe Application Forms~s soon as possible.

The next Intern ational SummerSchool will probably bete::'c. ::.r: Santiago de Chile in January-February 196~j.

"r

+++++.,-

- - - - - - - - - - - + ~ + + + + + ~ - + + ~ + + + + + + + + + + + + +